The location effect was reversed with a black background.
(A) The MED had an insignificant location dependence (p = 0.6259). The preferred location for white bar fixation was 10° above the flies. (B) The fixation performance at 2 locations (−20°; 30°) was selected to show additional details of the dwelling time in the whole panorama. There were 3 peaks in the distribution of the dwelling time (around −180°, +45° and +135° to the bar, respectively) at the location θ = −20°; there were 2 peaks in the distribution of the dwelling time (around and opposite the bar, respectively) at the location θ = 10°. (C) Box plots of pseudo PI. There was some location dependence in the pseudo PIs, although this tendency was not significant (p = 0.1044). (D) There was an insignificant location dependence in the rDTabs between quadrants around or opposite the bar and the rest area (p = 0.5959). (E) There was no obvious location dependence in the rDFS between quadrants around or opposite the bar and the rest area (p = 0.6124). These differences were quite small. The data in (A&B) are given as the mean±SEM; the data in (C–E) are shown by box plot; the red crosses represent outliers; the p values were calculated by the Kruskal-Wallis test; the y-axes of certain charts are truncated for compactness.